K-Means Based Matching Algorithm for Multi-Resolution Feature Descriptors
|
|
- Opal Stevens
- 6 years ago
- Views:
Transcription
1 K-Means Based Matching Algorithm for Multi-Resolution Feature Descriptors Shao-Tzu Huang, Chen-Chien Hsu, Wei-Yen Wang International Science Index, Electrical and Computer Engineering waset.org/publication/ Abstract Matching high dimensional features between images is computationally expensive for exhaustive search approaches in computer vision. Although the dimension of the feature can be degraded by simplifying the prior knowledge of homography, matching accuracy may degrade as a tradeoff. In this paper, we present a feature matching method based on k-means algorithm that reduces the matching cost and matches the features between images instead of using a simplified geometric assumption. Experimental results show that the proposed method outperforms the previous linear exhaustive search approaches in terms of the inlier ratio of matched pairs. Keywords Feature matching, k-means clustering, scale invariant feature transform, linear exhaustive search. I. INTRODUCTION EATURE matching is widely used in computer vision Fapplication such as object recognition, image registration, and virtual reality. In general, the task of matching feature points between two images contains three essential steps: interesting point detection, feature description, and feature matching. First, interesting points are selected at the specific location in the image. Next, each of the interesting points is represented by a feature vector, namely, feature descriptor, which describes the geometric properties of the point with strong resistance to the transformation of homography matrix. Finally, features between two images are matched by comparing the differences between each of the feature vectors. There are several well-known features detection methods [], such as Scale Invariant Feature Transform (SIFT) [2], Speeded-Up Robust Feature (SURF) [3], and Principal Components Analysis (PCA-SIFT) [4]. In practice, the dimension of feature vector depends on the amount of information that it used to describe the interest points distinctively. In the other words, feature vectors have high dimensionality for the descriptor design purpose of accuracy and robustness. Hence, the task of feature matching is to identify the most similar matches between different highdimensional vectors which are the most computationally expensive part. Given that exhaustive search is an intuitive and well-known approach for feature matching, the main challenge is that it requires a great amount of computation cost for high Shao-Tzu Huang is with the National Taiwan Normal University, Graduated Institution of Electrical Engineering Taipei, Taiwan ( glnhwng@hotmail.com). Chen-Chien Hsu and Wei-Yen Wang are with the National Taiwan Normal University, Department of Electrical Engineering Taipei, Taiwan ( jhsu@ntnu.edu.tw, wywang@ntnu.edu.tw). dimension feature matching. For example, if we have to match k dimension feature vectors of SIFT descriptors between two images, the computation cost increases as the order of O(kMN), where M and N are the number of the features on each image. Moreover, the overfitting problem may seriously cause the matching inaccuracy that requires an iterative filtering process to handle. Therefore, it is not applicable to real-time tasks which are essential for modern computer vision application. The present feature matching methods can be separated into two categories: distance metric and space partition. The former estimates the distance in feature vector space, and then, each of the two features in the different images which contain the shortest distance is formed as a feature pair. The exhaustive search method is an example and is improved by a considerable amount of literature for decades. Song et al. [5], [6] speed up the searching process by using a norm-sorted database. However, in real-time applications, the input data are non-sorted that need an additional computation cost of the sorting process before the use. Tsai et al. [7] apply Multiresolution Candidate Elimination (MRCE) technique to deal with such problem by simply the prior knowledge of geometric transform between the two images, and hence degrade the matching accuracy as a tradeoff. The space partition method creates a tree structure to efficiently speed up the search of the nearest neighbors. Using KD-tree for matching [8], is a typical example of the space partition method. Silpa-Anan and Hartley [9] use multiple KD-trees from the same data set to improve the search performance. Muja and Lowe [0] present the priority search k-means tree algorithm to approximate nearest neighbors. However, all of the space partition approaches require additional memory for storing the tree structures. Furthermore, the size of memory will set limit on the matching performance which is not desirable. According to the pros and cons of the feature matching approaches, we are motivated to develop a feature matching method that decreases the amount of redundant matching of exhaustive search approaches. Inspired by [7], our design reduces a great amount of redundant matching and hence overall degrades the computation cost of linear exhaustive search (LES) approaches. Furthermore, our matching method is designed without using a simplifying the prior knowledge of homography between the different images. Therefore, we leverage the applicability and the overall matching quality of LES approaches. 903
2 Start Construct L- norm-based pyramid K-means clustering Cluster matching Calculate D min of each cluster pair Pyramidbased descriptor matching End K-means cluster matching Pyramid-based descriptor matching Fig. The flow chart of proposed method International Science Index, Electrical and Computer Engineering waset.org/publication/ II. METHOD The proposed method reduces the computation cost of feature matching by using k-means clustering. It mainly contains three parts: ) construction of L norm based pyramid, 2) k-means cluster matching, and 3) pyramid-based descriptor matching. The details of each step are demonstrated in Fig. and explored in this section. A. Construction of L - Norm-Based Pyramid As described in the previous section, we target on reducing the computational cost of feature matching. Under the same scene, two images I and I 2 are captured at different camera positions. We utilize an existing algorithm to detect interest points in the image and represent each of the points with 2 l dimension vector. Inspired by [5]-[7], a pyramid structure is constructed on the feature vector domain for each interest point in the two images. Each layer of the pyramid contains one dimension information. From bottom to top of the pyramid, the element of layer i is computed by () i ( i) ( i) k 2k 2k u u u, k 2 i () where u ( i ) k is the k-th element of layer i. In the other words, there is only one element in the top layer of the pyramid. The whole pyramid is constructed without sorting process because of time-consuming concern. g i ( u) g ( ) i u / g ( u) 2 g ( u) Fig. 2 L -norm-based pyramid structure B. k-means Cluster Matching To reduce the computation cost of feature matching, the feature points in the image are clustered by using k-means algorithm []-[3]. That is, given the position vector x of each feature point, we have to solve the objective function g ( u) 0 k arg min x cj (2) S j xsj where c j is the position vector of the center point of j-th cluster S j. The center point c i is stored in the cluster matching process between the two images. We compute the Euclidian distance, and hence each of the two closest clusters at different image forms a cluster pair. The number of cluster pairs depends on the parameter k. C. Pyramid-Based Descriptor Matching The final step of the proposed method is a matching process between each two feature-based pyramid in different images. As mentioned in the previous section, our method seeks to reduce the amount of redundant feature matching. In practice, only the features in the same cluster pair are used for the match. For each cluster pair S j and S j2 on image I and I 2 respectively, we estimate the average Manhattan distance of the cluster pairs D g () u gˆ () u (3) min MN 0 0 g0( u) Sj, gˆ 0( u) Sj2 where M and N are the number of feature points in cluster S j and S j2, respectively. After that, the matching process begins from the top layers of each two pyramids. Fig. 3 The matching performance using various parameter k of k- means clustering. The blue line represents the average computation time. The green dashed line represents the average inlier ratio of matched pairs If the Manhattan distance between the elements of the top layer is smaller than D min, we will repeat the same matching 2 904
3 International Science Index, Electrical and Computer Engineering waset.org/publication/ process in the next layer of two pyramids until reaching the bottom layer of the pyramid. Otherwise, we consider the two features as unrelated feature points. Therefore, any matched feature pairs have to satisfy the condition that each of the Manhattan distance between the elements in the same layer of pyramids is smaller than D min. III. EXPERIMENT An experiment is conducted to evaluate our method. The input of our simulation tests is synthetic scenes from the benchmark of Nister and Stewenius [4] and Mikolajczyk amd Schmid [5]. We utilize SIFT algorithm [2], [8] to locate the feature points in different images and represent each of them with a 28-dimensional vector. Then, we apply our feature matching method to obtain matched feature pairs between different images. To evaluate the matching performance, the matched feature pairs are classified into inlier and outlier pairs according to the homography matrix []. In practice, we input the position data of matched feature pairs to the OpenCV findhomography function with Random Sample Consensus (RANSAC) method to handle the classification. Then, we compute the inlier ratio by R W / W (4) inlier where W inlier and W all are the number of inliers and the total feature pairs, respectively. Due to the random property of k- means algorithm, we run the method 0 times and compute the average inlier ratio. The computation time of the matching process is also considered in our test. TABLE I PERFORMANCE COMPARISON OF 600 SCENES Matching method LES MRCE [7] Proposed method Average Inlier ratio Average Time (sec.) Average Inlier ratio Average Time (sec.) Average Inlier ratio Average Time (sec.) Test scene TABLE II PERFORMANCE COMPARISON OF FIVE SELECTED SCENES Matching method LES MRCE [7] Proposed method Inlier ratio Time (sec.) Inlier ratio Time (sec.) Inlier ratio Time (sec.) Fig. 4 (a) Fig. 4 (b) Fig. 4 (c) Fig. 4 (d) Fig. 4 (e) To determine the parameter k of k-means clustering, we repeat the experiment with various k setting using 600 test scenes from the benchmark dataset. The experiment results are shown in Fig. 3. We can see that the computation time and the average inlier ratio of matched pairs increase as k increases from 3 to 6. To achieve the balance of joint matching performance, k should locate at the point before the crossing section of the two lines in Fig. 3. Therefore, we suggest setting k to be 4 for a desirable joint matching performance of both computation time and average inlier ratio of matched pairs. We compare the matching performance of our method with LES and MRCE [7] with another 600 test scenes. The comparison is shown in Table I. We can see that our method outperforms LES and MRCE in the comparison of the average inlier ratio of matched pairs. Although the computation time of this method is slower than MRCE, the average ratio of inlier feature pairs is higher than that of MRCE. This is because our k-means clustering process suppresses the impact of D min inaccurate estimate. As described in the previous section, the computation of D min is required for the matching process. The physical meaning of D min is a generalized distance of the geometric transform between different images. If wrong matching feature pairs exist in the computation, which are extreme values of the input, estimation error of D min occurs due to the averaging operator. MRCE does not contain clustering process before the D min computation. Hence, it simplifies the prior knowledge of D min as a fixed distance of the geometric transform for every matched pair in that MRCE suffers from the inaccurate D min problem. This causes a serious impact when the two matching images are captured in the same scene but with different illuminations or blurriness. To elaborate, we empirically select five typical scenes from the 600 test scenes as shown in Fig. 4. The first two image pairs are images captured in different camera positions, and the third pair is in different zoom settings. The fourth and fifth image pairs are the same images with different blurriness and illumination, respectively. The experimental results of the five scenes are shown in Table II. In Table II, we can see that the inlier ratio of MRCE is the lowest in the comparison of test scene Figs. 4 (d) and (e) due to inaccurate D min problem. On the contrary, the k-means clustering process of our method reduces some sort of wrong feature matches from the computation and hence suppresses the impact of wrong matching pairs. In this circumstance, the accuracy of the D min estimate will be improved. As a result, the introduced method will effectively reduce the redundant matching and it improves the matching accuracy. all 905
4 (a) (b) (c) International Science Index, Electrical and Computer Engineering waset.org/publication/ (d) (e) Fig. 4 The comparison of matched feature pairs between LES (left), and MRCE [7] (middle), and the proposed method (right). The green lines and red line represents inlier and outlier feature pairs, respectively. (a) and (b) are the affine transformed images, (c) are the scale changed images, (d) are the blurred images, and (e) are the illumination changed images IV. CONCLUSION We have described a feature matching method based on combining L -norm based pyramid and k-means clustering. Our method prevails over the conventional brute force method and previous linear exhaustive approaches in the average ratio of inlier matched feature pairs. The method lays the foundation of the robust feature matching without using a simplified prior knowledge of homography and additional memory needs. ACKNOWLEDGMENT This research is partially supported by the Center of Learning Technology for Chinese and Aim for the Top University Project of National Taiwan Normal University (NTNU), sponsored by the Ministry of Education, Taiwan, R.O.C. and Ministry of Science and Technology, Taiwan, R.O.C. under Grant no. MOST E REFERENCES [] L. Juan and O. Gwun. A Comparison of SIFT, PCA-SIFT and SURF, International Journal of Image Processing, Vol. 65, pp , [2] D. G. Lowe. Distinctive image features from scale-invariant keypoints, International journal of computer vision, vol. 60, No. 2, pp. 9-0, [3] H. Bay, A. Ess, T. Tuytelaars, and L. Van Gool. Speeded-up robust features (SURF), Computer vision and image understanding, vol. 0, No. 3, pp , [4] Y. Ke and R. Sukthankar. PCA-SIFT a more distinctive representation for local image descriptors, Proc. International Conference on Computer Vision and Pattern Recognition, 2004, Vol. 2, pp [5] B. C. Song and J. B. Ra, Multiresolution descriptor matching algorithm for fast exhaustive search in norm-sorted databases, Journal of Electronic Imaging, vol. 4, No.4, pp , [6] B. C. Song, M. J. Kim, and J. B. Ra. A fast multiresolution feature matching algorithm for exhaustive search in large image databases, IEEE Transactions on Circuits and Systems for Video Technology, vol., No. 5, pp , 200. [7] C.-Y. Tsai, A.-H. Tsao, and C.-W. Wang. "Real-time feature descriptor matching via a multi-resolution exhaustive search method," Journal of Software, vol. 8, no. 9, pp , 203. [8] D. G. Lowe. Object recognition from local scale-invariant features, In Computer vision, The proceedings of the seventh IEEE international conference on IEEE, 999, Vol. 2, pp [9] C. Silpa-Anan and R. Hartley. Optimised KD-trees for fast image descriptor matching, In Proc. Computer Vision and Pattern Recognition, CVPR IEEE Conference on IEEE, June, 2008, pp [0] M. Muja and D. G. Lowe. "Scalable nearest neighbor algorithms for high dimensional data," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 36, No., pp , 204. [] J. A. Hartigan and M. A. Wong, "Algorithm AS 36: A k-means clustering algorithm." Journal of the Royal Statistical Society. Series C (Applied Statistics), Vol. 28, No., pp , 979. [2] K. Wagstaff, C. Cardie, S. Rogers, and S. Schrödl. Constrained k- means clustering with background knowledge, In ICML, June, 200, Vol., pp [3] V. Hautamäki, S. Cherednichenko, I. Kärkkäinen, T. Kinnunen, and P. Fränti. Improving k-means by outlier removal, In Scandinavian Conference on Image Analysis. Springer Berlin Heidelberg, 2005, pp. 906
5 [4] D. Nister and H. Stewenius, Scalable recognition with a vocabulary tree, Proc. International Conference on Computer Vision and Pattern Recognition, 2006, Vol. 2, pp [5] K. Mikolajczyk and C. Schmid, "A performance evaluation of local descriptors," IEEE transactions on pattern analysis and machine intelligence, vol27, No.0, pp , International Science Index, Electrical and Computer Engineering waset.org/publication/
A Comparison of SIFT, PCA-SIFT and SURF
A Comparison of SIFT, PCA-SIFT and SURF Luo Juan Computer Graphics Lab, Chonbuk National University, Jeonju 561-756, South Korea qiuhehappy@hotmail.com Oubong Gwun Computer Graphics Lab, Chonbuk National
More informationIII. VERVIEW OF THE METHODS
An Analytical Study of SIFT and SURF in Image Registration Vivek Kumar Gupta, Kanchan Cecil Department of Electronics & Telecommunication, Jabalpur engineering college, Jabalpur, India comparing the distance
More informationSpecular 3D Object Tracking by View Generative Learning
Specular 3D Object Tracking by View Generative Learning Yukiko Shinozuka, Francois de Sorbier and Hideo Saito Keio University 3-14-1 Hiyoshi, Kohoku-ku 223-8522 Yokohama, Japan shinozuka@hvrl.ics.keio.ac.jp
More informationSURF applied in Panorama Image Stitching
Image Processing Theory, Tools and Applications SURF applied in Panorama Image Stitching Luo Juan 1, Oubong Gwun 2 Computer Graphics Lab, Computer Science & Computer Engineering, Chonbuk National University,
More informationLOCAL AND GLOBAL DESCRIPTORS FOR PLACE RECOGNITION IN ROBOTICS
8th International DAAAM Baltic Conference "INDUSTRIAL ENGINEERING - 19-21 April 2012, Tallinn, Estonia LOCAL AND GLOBAL DESCRIPTORS FOR PLACE RECOGNITION IN ROBOTICS Shvarts, D. & Tamre, M. Abstract: The
More informationRobot localization method based on visual features and their geometric relationship
, pp.46-50 http://dx.doi.org/10.14257/astl.2015.85.11 Robot localization method based on visual features and their geometric relationship Sangyun Lee 1, Changkyung Eem 2, and Hyunki Hong 3 1 Department
More informationImplementation and Comparison of Feature Detection Methods in Image Mosaicing
IOSR Journal of Electronics and Communication Engineering (IOSR-JECE) e-issn: 2278-2834,p-ISSN: 2278-8735 PP 07-11 www.iosrjournals.org Implementation and Comparison of Feature Detection Methods in Image
More informationIMAGE RETRIEVAL USING VLAD WITH MULTIPLE FEATURES
IMAGE RETRIEVAL USING VLAD WITH MULTIPLE FEATURES Pin-Syuan Huang, Jing-Yi Tsai, Yu-Fang Wang, and Chun-Yi Tsai Department of Computer Science and Information Engineering, National Taitung University,
More informationFeature Based Registration - Image Alignment
Feature Based Registration - Image Alignment Image Registration Image registration is the process of estimating an optimal transformation between two or more images. Many slides from Alexei Efros http://graphics.cs.cmu.edu/courses/15-463/2007_fall/463.html
More informationFeature Detection and Matching
and Matching CS4243 Computer Vision and Pattern Recognition Leow Wee Kheng Department of Computer Science School of Computing National University of Singapore Leow Wee Kheng (CS4243) Camera Models 1 /
More informationA Novel Extreme Point Selection Algorithm in SIFT
A Novel Extreme Point Selection Algorithm in SIFT Ding Zuchun School of Electronic and Communication, South China University of Technolog Guangzhou, China zucding@gmail.com Abstract. This paper proposes
More informationEvaluation and comparison of interest points/regions
Introduction Evaluation and comparison of interest points/regions Quantitative evaluation of interest point/region detectors points / regions at the same relative location and area Repeatability rate :
More informationDeterminant of homography-matrix-based multiple-object recognition
Determinant of homography-matrix-based multiple-object recognition 1 Nagachetan Bangalore, Madhu Kiran, Anil Suryaprakash Visio Ingenii Limited F2-F3 Maxet House Liverpool Road Luton, LU1 1RS United Kingdom
More informationImprovement of SURF Feature Image Registration Algorithm Based on Cluster Analysis
Sensors & Transducers 2014 by IFSA Publishing, S. L. http://www.sensorsportal.com Improvement of SURF Feature Image Registration Algorithm Based on Cluster Analysis 1 Xulin LONG, 1,* Qiang CHEN, 2 Xiaoya
More informationObject Recognition with Invariant Features
Object Recognition with Invariant Features Definition: Identify objects or scenes and determine their pose and model parameters Applications Industrial automation and inspection Mobile robots, toys, user
More informationA Rapid Automatic Image Registration Method Based on Improved SIFT
Available online at www.sciencedirect.com Procedia Environmental Sciences 11 (2011) 85 91 A Rapid Automatic Image Registration Method Based on Improved SIFT Zhu Hongbo, Xu Xuejun, Wang Jing, Chen Xuesong,
More informationA NEW FEATURE BASED IMAGE REGISTRATION ALGORITHM INTRODUCTION
A NEW FEATURE BASED IMAGE REGISTRATION ALGORITHM Karthik Krish Stuart Heinrich Wesley E. Snyder Halil Cakir Siamak Khorram North Carolina State University Raleigh, 27695 kkrish@ncsu.edu sbheinri@ncsu.edu
More informationEnsemble of Bayesian Filters for Loop Closure Detection
Ensemble of Bayesian Filters for Loop Closure Detection Mohammad Omar Salameh, Azizi Abdullah, Shahnorbanun Sahran Pattern Recognition Research Group Center for Artificial Intelligence Faculty of Information
More informationTensor Decomposition of Dense SIFT Descriptors in Object Recognition
Tensor Decomposition of Dense SIFT Descriptors in Object Recognition Tan Vo 1 and Dat Tran 1 and Wanli Ma 1 1- Faculty of Education, Science, Technology and Mathematics University of Canberra, Australia
More informationDetecting Printed and Handwritten Partial Copies of Line Drawings Embedded in Complex Backgrounds
9 1th International Conference on Document Analysis and Recognition Detecting Printed and Handwritten Partial Copies of Line Drawings Embedded in Complex Backgrounds Weihan Sun, Koichi Kise Graduate School
More informationVideo Processing for Judicial Applications
Video Processing for Judicial Applications Konstantinos Avgerinakis, Alexia Briassouli, Ioannis Kompatsiaris Informatics and Telematics Institute, Centre for Research and Technology, Hellas Thessaloniki,
More informationA Novel Real-Time Feature Matching Scheme
Sensors & Transducers, Vol. 165, Issue, February 01, pp. 17-11 Sensors & Transducers 01 by IFSA Publishing, S. L. http://www.sensorsportal.com A Novel Real-Time Feature Matching Scheme Ying Liu, * Hongbo
More informationFast Image Matching Using Multi-level Texture Descriptor
Fast Image Matching Using Multi-level Texture Descriptor Hui-Fuang Ng *, Chih-Yang Lin #, and Tatenda Muindisi * Department of Computer Science, Universiti Tunku Abdul Rahman, Malaysia. E-mail: nghf@utar.edu.my
More informationA System of Image Matching and 3D Reconstruction
A System of Image Matching and 3D Reconstruction CS231A Project Report 1. Introduction Xianfeng Rui Given thousands of unordered images of photos with a variety of scenes in your gallery, you will find
More informationImprovements to Uncalibrated Feature-Based Stereo Matching for Document Images by using Text-Line Segmentation
Improvements to Uncalibrated Feature-Based Stereo Matching for Document Images by using Text-Line Segmentation Muhammad Zeshan Afzal, Martin Krämer, Syed Saqib Bukhari, Faisal Shafait and Thomas M. Breuel
More informationStereoscopic Images Generation By Monocular Camera
Stereoscopic Images Generation By Monocular Camera Swapnil Lonare M. tech Student Department of Electronics Engineering (Communication) Abha Gaikwad - Patil College of Engineering. Nagpur, India 440016
More informationGeneralized RANSAC framework for relaxed correspondence problems
Generalized RANSAC framework for relaxed correspondence problems Wei Zhang and Jana Košecká Department of Computer Science George Mason University Fairfax, VA 22030 {wzhang2,kosecka}@cs.gmu.edu Abstract
More informationToward Part-based Document Image Decoding
2012 10th IAPR International Workshop on Document Analysis Systems Toward Part-based Document Image Decoding Wang Song, Seiichi Uchida Kyushu University, Fukuoka, Japan wangsong@human.ait.kyushu-u.ac.jp,
More informationCollecting outdoor datasets for benchmarking vision based robot localization
Collecting outdoor datasets for benchmarking vision based robot localization Emanuele Frontoni*, Andrea Ascani, Adriano Mancini, Primo Zingaretti Department of Ingegneria Infromatica, Gestionale e dell
More informationKey properties of local features
Key properties of local features Locality, robust against occlusions Must be highly distinctive, a good feature should allow for correct object identification with low probability of mismatch Easy to etract
More informationApplying Multiple KD-Trees in High Dimensional Nearest Neighbor Searching
Applying Multiple KD-Trees in High Dimensional Nearest Neighbor Searching Shwu-Huey Yen, Chao-Yu Shih, Tai-Kuang Li, Hsiao-Wei Chang Abstract Feature matching plays a key role in many image processing
More informationViewpoint Invariant Features from Single Images Using 3D Geometry
Viewpoint Invariant Features from Single Images Using 3D Geometry Yanpeng Cao and John McDonald Department of Computer Science National University of Ireland, Maynooth, Ireland {y.cao,johnmcd}@cs.nuim.ie
More informationA Comparison of SIFT and SURF
A Comparison of SIFT and SURF P M Panchal 1, S R Panchal 2, S K Shah 3 PG Student, Department of Electronics & Communication Engineering, SVIT, Vasad-388306, India 1 Research Scholar, Department of Electronics
More informationA Method to Eliminate Wrongly Matched Points for Image Matching
2017 2nd International Seminar on Applied Physics, Optoelectronics and Photonics (APOP 2017) ISBN: 978-1-60595-522-3 A Method to Eliminate Wrongly Matched Points for Image Matching Xiao-fei AI * ABSTRACT
More informationSemantic Kernels Binarized A Feature Descriptor for Fast and Robust Matching
2011 Conference for Visual Media Production Semantic Kernels Binarized A Feature Descriptor for Fast and Robust Matching Frederik Zilly 1, Christian Riechert 1, Peter Eisert 1,2, Peter Kauff 1 1 Fraunhofer
More informationAutomatic Image Alignment
Automatic Image Alignment Mike Nese with a lot of slides stolen from Steve Seitz and Rick Szeliski 15-463: Computational Photography Alexei Efros, CMU, Fall 2010 Live Homography DEMO Check out panoramio.com
More informationFuzzy based Multiple Dictionary Bag of Words for Image Classification
Available online at www.sciencedirect.com Procedia Engineering 38 (2012 ) 2196 2206 International Conference on Modeling Optimisation and Computing Fuzzy based Multiple Dictionary Bag of Words for Image
More informationA Hybrid Feature Extractor using Fast Hessian Detector and SIFT
Technologies 2015, 3, 103-110; doi:10.3390/technologies3020103 OPEN ACCESS technologies ISSN 2227-7080 www.mdpi.com/journal/technologies Article A Hybrid Feature Extractor using Fast Hessian Detector and
More informationCodebook Graph Coding of Descriptors
Int'l Conf. Par. and Dist. Proc. Tech. and Appl. PDPTA'5 3 Codebook Graph Coding of Descriptors Tetsuya Yoshida and Yuu Yamada Graduate School of Humanities and Science, Nara Women s University, Nara,
More informationSIFT: SCALE INVARIANT FEATURE TRANSFORM SURF: SPEEDED UP ROBUST FEATURES BASHAR ALSADIK EOS DEPT. TOPMAP M13 3D GEOINFORMATION FROM IMAGES 2014
SIFT: SCALE INVARIANT FEATURE TRANSFORM SURF: SPEEDED UP ROBUST FEATURES BASHAR ALSADIK EOS DEPT. TOPMAP M13 3D GEOINFORMATION FROM IMAGES 2014 SIFT SIFT: Scale Invariant Feature Transform; transform image
More informationKEY FRAMES EXTRACTION USING GRAPH MODULARITY CLUSTERING FOR EFFICIENT VIDEO SUMMARIZATION
KEY FRAMES EXTRACTION USING GRAPH MODULARITY CLUSTERING FOR EFFICIENT VIDEO SUMMARIZATION Hana Gharbi, Sahbi Bahroun, Mohamed Massaoudi and Ezzeddine Zagrouba Laboratoire LIMTIC, Institut Supérieur d'informatqiue,
More informationScale Invariant Feature Transform
Scale Invariant Feature Transform Why do we care about matching features? Camera calibration Stereo Tracking/SFM Image moiaicing Object/activity Recognition Objection representation and recognition Image
More informationLocal features and image matching. Prof. Xin Yang HUST
Local features and image matching Prof. Xin Yang HUST Last time RANSAC for robust geometric transformation estimation Translation, Affine, Homography Image warping Given a 2D transformation T and a source
More informationImage Features: Local Descriptors. Sanja Fidler CSC420: Intro to Image Understanding 1/ 58
Image Features: Local Descriptors Sanja Fidler CSC420: Intro to Image Understanding 1/ 58 [Source: K. Grauman] Sanja Fidler CSC420: Intro to Image Understanding 2/ 58 Local Features Detection: Identify
More informationAutomatic Image Alignment
Automatic Image Alignment with a lot of slides stolen from Steve Seitz and Rick Szeliski Mike Nese CS194: Image Manipulation & Computational Photography Alexei Efros, UC Berkeley, Fall 2018 Live Homography
More informationarxiv: v1 [cs.cv] 28 Sep 2018
Camera Pose Estimation from Sequence of Calibrated Images arxiv:1809.11066v1 [cs.cv] 28 Sep 2018 Jacek Komorowski 1 and Przemyslaw Rokita 2 1 Maria Curie-Sklodowska University, Institute of Computer Science,
More informationSupplementary Material for Ensemble Diffusion for Retrieval
Supplementary Material for Ensemble Diffusion for Retrieval Song Bai 1, Zhichao Zhou 1, Jingdong Wang, Xiang Bai 1, Longin Jan Latecki 3, Qi Tian 4 1 Huazhong University of Science and Technology, Microsoft
More informationLight-Weight Spatial Distribution Embedding of Adjacent Features for Image Search
Light-Weight Spatial Distribution Embedding of Adjacent Features for Image Search Yan Zhang 1,2, Yao Zhao 1,2, Shikui Wei 3( ), and Zhenfeng Zhu 1,2 1 Institute of Information Science, Beijing Jiaotong
More informationA Comparison and Matching Point Extraction of SIFT and ISIFT
A Comparison and Matching Point Extraction of SIFT and ISIFT A. Swapna A. Geetha Devi M.Tech Scholar, PVPSIT, Vijayawada Associate Professor, PVPSIT, Vijayawada bswapna.naveen@gmail.com geetha.agd@gmail.com
More informationSpeeding up the Detection of Line Drawings Using a Hash Table
Speeding up the Detection of Line Drawings Using a Hash Table Weihan Sun, Koichi Kise 2 Graduate School of Engineering, Osaka Prefecture University, Japan sunweihan@m.cs.osakafu-u.ac.jp, 2 kise@cs.osakafu-u.ac.jp
More informationThe SIFT (Scale Invariant Feature
The SIFT (Scale Invariant Feature Transform) Detector and Descriptor developed by David Lowe University of British Columbia Initial paper ICCV 1999 Newer journal paper IJCV 2004 Review: Matt Brown s Canonical
More informationFeature-based methods for image matching
Feature-based methods for image matching Bag of Visual Words approach Feature descriptors SIFT descriptor SURF descriptor Geometric consistency check Vocabulary tree Digital Image Processing: Bernd Girod,
More informationFast Natural Feature Tracking for Mobile Augmented Reality Applications
Fast Natural Feature Tracking for Mobile Augmented Reality Applications Jong-Seung Park 1, Byeong-Jo Bae 2, and Ramesh Jain 3 1 Dept. of Computer Science & Eng., University of Incheon, Korea 2 Hyundai
More informationSEARCH BY MOBILE IMAGE BASED ON VISUAL AND SPATIAL CONSISTENCY. Xianglong Liu, Yihua Lou, Adams Wei Yu, Bo Lang
SEARCH BY MOBILE IMAGE BASED ON VISUAL AND SPATIAL CONSISTENCY Xianglong Liu, Yihua Lou, Adams Wei Yu, Bo Lang State Key Laboratory of Software Development Environment Beihang University, Beijing 100191,
More informationObject and Action Detection from a Single Example
Object and Action Detection from a Single Example Peyman Milanfar* EE Department University of California, Santa Cruz *Joint work with Hae Jong Seo AFOSR Program Review, June 4-5, 29 Take a look at this:
More informationScale Invariant Feature Transform
Why do we care about matching features? Scale Invariant Feature Transform Camera calibration Stereo Tracking/SFM Image moiaicing Object/activity Recognition Objection representation and recognition Automatic
More informationImproving Recognition through Object Sub-categorization
Improving Recognition through Object Sub-categorization Al Mansur and Yoshinori Kuno Graduate School of Science and Engineering, Saitama University, 255 Shimo-Okubo, Sakura-ku, Saitama-shi, Saitama 338-8570,
More informationAutomatic Image Alignment (feature-based)
Automatic Image Alignment (feature-based) Mike Nese with a lot of slides stolen from Steve Seitz and Rick Szeliski 15-463: Computational Photography Alexei Efros, CMU, Fall 2006 Today s lecture Feature
More informationAn Evaluation of Volumetric Interest Points
An Evaluation of Volumetric Interest Points Tsz-Ho YU Oliver WOODFORD Roberto CIPOLLA Machine Intelligence Lab Department of Engineering, University of Cambridge About this project We conducted the first
More informationarxiv: v1 [cs.cv] 28 Sep 2018
Extrinsic camera calibration method and its performance evaluation Jacek Komorowski 1 and Przemyslaw Rokita 2 arxiv:1809.11073v1 [cs.cv] 28 Sep 2018 1 Maria Curie Sklodowska University Lublin, Poland jacek.komorowski@gmail.com
More informationYudistira Pictures; Universitas Brawijaya
Evaluation of Feature Detector-Descriptor for Real Object Matching under Various Conditions of Ilumination and Affine Transformation Novanto Yudistira1, Achmad Ridok2, Moch Ali Fauzi3 1) Yudistira Pictures;
More informationCHAPTER 1 INTRODUCTION
1 CHAPTER 1 INTRODUCTION Table of Contents Page No. 1 INTRODUCTION 1.1 Problem overview 2 1.2 Research objective 3 1.3 Thesis outline 7 2 1. INTRODUCTION 1.1 PROBLEM OVERVIEW The process of mapping and
More informationCPPP/UFMS at ImageCLEF 2014: Robot Vision Task
CPPP/UFMS at ImageCLEF 2014: Robot Vision Task Rodrigo de Carvalho Gomes, Lucas Correia Ribas, Amaury Antônio de Castro Junior, Wesley Nunes Gonçalves Federal University of Mato Grosso do Sul - Ponta Porã
More informationMirror Match: Reliable Feature Point Matching Without Geometric Constraints
Mirror Match: Reliable Feature Point Matching Without Geometric Constraints Jonas Toft Arnfred, Stefan Winkler, Sabine Süsstrunk Advanced Digital Sciences Center (ADSC), University of Illinois at Urbana-Champaign
More informationOn-line Document Registering and Retrieving System for AR Annotation Overlay
On-line Document Registering and Retrieving System for AR Annotation Overlay Hideaki Uchiyama, Julien Pilet and Hideo Saito Keio University 3-14-1 Hiyoshi, Kohoku-ku Yokohama, Japan {uchiyama,julien,saito}@hvrl.ics.keio.ac.jp
More informationAutomatic estimation of the inlier threshold in robust multiple structures fitting.
Automatic estimation of the inlier threshold in robust multiple structures fitting. Roberto Toldo and Andrea Fusiello Dipartimento di Informatica, Università di Verona Strada Le Grazie, 3734 Verona, Italy
More informationFace Recognition using SURF Features and SVM Classifier
International Journal of Electronics Engineering Research. ISSN 0975-6450 Volume 8, Number 1 (016) pp. 1-8 Research India Publications http://www.ripublication.com Face Recognition using SURF Features
More informationPerformance Evaluation of Scale-Interpolated Hessian-Laplace and Haar Descriptors for Feature Matching
Performance Evaluation of Scale-Interpolated Hessian-Laplace and Haar Descriptors for Feature Matching Akshay Bhatia, Robert Laganière School of Information Technology and Engineering University of Ottawa
More information3. Working Implementation of Search Engine 3.1 Working Of Search Engine:
Image Search Engine Using SIFT Algorithm Joshi Parag #, Shinde Saiprasad Prakash 1, Panse Mihir 2, Vaidya Omkar 3, # Assistance Professor and 1, 2 &3 Students Department:- Computer Engineering, Rajendra
More informationVideo annotation based on adaptive annular spatial partition scheme
Video annotation based on adaptive annular spatial partition scheme Guiguang Ding a), Lu Zhang, and Xiaoxu Li Key Laboratory for Information System Security, Ministry of Education, Tsinghua National Laboratory
More information3D Environment Reconstruction
3D Environment Reconstruction Using Modified Color ICP Algorithm by Fusion of a Camera and a 3D Laser Range Finder The 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems October 11-15,
More informationKeypoint-based Recognition and Object Search
03/08/11 Keypoint-based Recognition and Object Search Computer Vision CS 543 / ECE 549 University of Illinois Derek Hoiem Notices I m having trouble connecting to the web server, so can t post lecture
More informationMotion Estimation and Optical Flow Tracking
Image Matching Image Retrieval Object Recognition Motion Estimation and Optical Flow Tracking Example: Mosiacing (Panorama) M. Brown and D. G. Lowe. Recognising Panoramas. ICCV 2003 Example 3D Reconstruction
More information3D model search and pose estimation from single images using VIP features
3D model search and pose estimation from single images using VIP features Changchang Wu 2, Friedrich Fraundorfer 1, 1 Department of Computer Science ETH Zurich, Switzerland {fraundorfer, marc.pollefeys}@inf.ethz.ch
More informationFeature Detection. Raul Queiroz Feitosa. 3/30/2017 Feature Detection 1
Feature Detection Raul Queiroz Feitosa 3/30/2017 Feature Detection 1 Objetive This chapter discusses the correspondence problem and presents approaches to solve it. 3/30/2017 Feature Detection 2 Outline
More informationSemi-Supervised Clustering with Partial Background Information
Semi-Supervised Clustering with Partial Background Information Jing Gao Pang-Ning Tan Haibin Cheng Abstract Incorporating background knowledge into unsupervised clustering algorithms has been the subject
More informationInternational Journal of Advanced Research in Computer Science and Software Engineering
ISSN: 2277 128X International Journal of Advanced Research in Computer Science and Software Engineering Research Paper Available online at: Fingerprint Recognition using Robust Local Features Madhuri and
More informationVK Multimedia Information Systems
VK Multimedia Information Systems Mathias Lux, mlux@itec.uni-klu.ac.at Dienstags, 16.oo Uhr This work is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Agenda Evaluations
More informationInstance-level recognition
Instance-level recognition 1) Local invariant features 2) Matching and recognition with local features 3) Efficient visual search 4) Very large scale indexing Matching of descriptors Matching and 3D reconstruction
More informationA Systems View of Large- Scale 3D Reconstruction
Lecture 23: A Systems View of Large- Scale 3D Reconstruction Visual Computing Systems Goals and motivation Construct a detailed 3D model of the world from unstructured photographs (e.g., Flickr, Facebook)
More informationCSE 527: Introduction to Computer Vision
CSE 527: Introduction to Computer Vision Week 5 - Class 1: Matching, Stitching, Registration September 26th, 2017 ??? Recap Today Feature Matching Image Alignment Panoramas HW2! Feature Matches Feature
More informationImproved Coding for Image Feature Location Information
Improved Coding for Image Feature Location Information Sam S. Tsai, David Chen, Gabriel Takacs, Vijay Chandrasekhar Mina Makar, Radek Grzeszczuk, and Bernd Girod Department of Electrical Engineering, Stanford
More informationLarge Scale Image Retrieval
Large Scale Image Retrieval Ondřej Chum and Jiří Matas Center for Machine Perception Czech Technical University in Prague Features Affine invariant features Efficient descriptors Corresponding regions
More informationCS231A Section 6: Problem Set 3
CS231A Section 6: Problem Set 3 Kevin Wong Review 6 -! 1 11/09/2012 Announcements PS3 Due 2:15pm Tuesday, Nov 13 Extra Office Hours: Friday 6 8pm Huang Common Area, Basement Level. Review 6 -! 2 Topics
More informationInstance-level recognition part 2
Visual Recognition and Machine Learning Summer School Paris 2011 Instance-level recognition part 2 Josef Sivic http://www.di.ens.fr/~josef INRIA, WILLOW, ENS/INRIA/CNRS UMR 8548 Laboratoire d Informatique,
More informationAppearance-Based Place Recognition Using Whole-Image BRISK for Collaborative MultiRobot Localization
Appearance-Based Place Recognition Using Whole-Image BRISK for Collaborative MultiRobot Localization Jung H. Oh, Gyuho Eoh, and Beom H. Lee Electrical and Computer Engineering, Seoul National University,
More informationAn Angle Estimation to Landmarks for Autonomous Satellite Navigation
5th International Conference on Environment, Materials, Chemistry and Power Electronics (EMCPE 2016) An Angle Estimation to Landmarks for Autonomous Satellite Navigation Qing XUE a, Hongwen YANG, Jian
More informationComputer Vision for HCI. Topics of This Lecture
Computer Vision for HCI Interest Points Topics of This Lecture Local Invariant Features Motivation Requirements, Invariances Keypoint Localization Features from Accelerated Segment Test (FAST) Harris Shi-Tomasi
More informationObject Detection by Point Feature Matching using Matlab
Object Detection by Point Feature Matching using Matlab 1 Faishal Badsha, 2 Rafiqul Islam, 3,* Mohammad Farhad Bulbul 1 Department of Mathematics and Statistics, Bangladesh University of Business and Technology,
More informationAUTOMATIC MATCHING AND GEO-REFERENCING OF HISTORICAL AERIAL IMAGES
AUTOMATIC MATCHING AND GEO-REFERENCING OF HISTORICAL AERIAL IMAGES I-Wen Chen 1, Hou-Ren Chen 2, Yi-Hsing Tseng b Department of Geomatics, National Cheng Kung University, No.1, University Rd., Tainan City
More informationRobotic Grasping Based on Efficient Tracking and Visual Servoing using Local Feature Descriptors
INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING Vol. 13, No. 3, pp. 387-393 MARCH 2012 / 387 DOI: 10.1007/s12541-012-0049-8 Robotic Grasping Based on Efficient Tracking and Visual Servoing
More informationLocalisation using an image-map
Localisation using an image-map Chanop Silpa-Anan The Australian National University chanop@syseng.anu.edu.au Richard Hartley National ict Australia hartley@syseng.anu.edu.au Abstract This paper looks
More informationInstance-level recognition II.
Reconnaissance d objets et vision artificielle 2010 Instance-level recognition II. Josef Sivic http://www.di.ens.fr/~josef INRIA, WILLOW, ENS/INRIA/CNRS UMR 8548 Laboratoire d Informatique, Ecole Normale
More informationReal-time Vehicle Matching for Multi-camera Tunnel Surveillance
Real-time Vehicle Matching for Multi-camera Tunnel Surveillance Vedran Jelača, Jorge Oswaldo Niño Castañeda, Andrés Frías-Velázquez, Aleksandra Pižurica and Wilfried Philips ABSTRACT Tracking multiple
More informationTexture Classification Using an Invariant Texture Representation and a Tree Matching Kernel
99 Texture Classification Using an Invariant Texture Representation and a Tree Matching Kernel Somkid Soottitantawat 1 and Surapong Auwatanamongkol 2 1 School of Applied Statistics, National Institute
More informationP-SURF: A Robust Local Image Descriptor *
JOURNAL OF INFORMATION SCIENCE AND ENGINEERING 27, 2-25 (2) : A Robust Local Image Descriptor * CONGXIN LIU, JIE YANG AND HAI HUANG + Institute of Image Processing and Pattern Recognition Shanghai Jiao
More informationarxiv: v1 [cs.cv] 1 Jan 2019
Mapping Areas using Computer Vision Algorithms and Drones Bashar Alhafni Saulo Fernando Guedes Lays Cavalcante Ribeiro Juhyun Park Jeongkyu Lee University of Bridgeport. Bridgeport, CT, 06606. United States
More informationSIFT-Rank: Ordinal Description for Invariant Feature Correspondence
-Rank: Ordinal Description for Invariant Feature Correspondence Matthew Toews and William Wells III Harvard Medical School, Brigham and Women s Hospital {mt,sw}@bwh.harvard.edu Abstract This paper investigates
More informationRealtime Storefront Logo Recognition
Realtime Storefront Logo Recognition Frank Liu and Yanlin Chen {liuf, yanlinc}@stanford.edu Stanford University Abstract Storefront brand image (logo) recognition has many useful applications such as location
More informationStable Structure from Motion for Unordered Image Collections
Stable Structure from Motion for Unordered Image Collections Carl Olsson and Olof Enqvist Centre for Mathematical Sciences, Lund University, Sweden Abstract. We present a non-incremental approach to structure
More information